Open-weight LLM leaderboard · live

Local LLM leaderboard

Every open-weight model, ranked across what matters. Pick your hardware and the board re-ranks to what you can actually run.Verified May 27, 2026

All models

5 models

Rank for your hardwarecustom

#			↗	↗	↗	↗		Providers
01	Qwen 2.5 72B Instructqwen· stale	—	$327272.76	—	—	131k	Multi-GPU	1
02	DeepSeek R1 Distill Llama 70Bdeepseek· stale	—	$636363.71	—	—	131k	Multi-GPU	1
03	DeepSeek V3deepseek· stale	—	$290909.17	—	—	131k	Multi-GPU	1
04	Llama 3.1 70B Instructllama· stale	—	$363636.40	—	—	131k	Multi-GPU	1
05	Llama 3.1 8B Instructllama· stale	—	$1818.19	—	—	131k	24 GBlocal	1

Loading leaderboard…

Intelligence Index and hardware tier are derived estimates — see the methodology. Hardware VRAM is an estimate at the best available quantization.

all verified May 27, 2026